TikTok Trapped by Deepfakes of Taylor Swift and Rihanna

Celebrity Deepfakes: A New Weapon for Scammers on TikTok

AI-generated videos featuring celebrities like Taylor Swift and Rihanna are being used by scammers to promote dubious services on TikTok, according to authentication company Copyleaks.

These advertisements exploit images of celebrities in familiar contexts such as red carpet interviews, podcasts, or talk shows. They often manipulate real footage using AI to create an illusion of legitimacy. Many of these videos promote reward programs, claiming that users can earn money by watching TikTok content and sharing their opinions. Although the official TikTok brand appears in some ads, users are actually redirected to third-party services that request personal information.

AI Avatars to Deceive Users

In one advertisement, a realistic AI avatar of Taylor Swift encourages users to sign up for a feature called TikTok Pay. Another video features a fake Rihanna stating, "you’re literally watching content and giving your opinion."

These examples highlight the challenges faced by social platforms in combating the proliferation of convincing deepfakes, which are becoming a daily issue for users. TikTok is not alone in facing this threat. Reports indicate that users on Meta platforms, including Instagram and Facebook, are exposed to billions of fraudulent ads every day. Meta's oversight board has acknowledged the existence of a deepfake problem. YouTube, for its part, claims to be investing heavily to combat fraudulent ads featuring celebrities.

Celebrities Strike Back

In response to this threat, celebrities are seeking new ways to protect themselves. Last week, Taylor Swift filed new trademark applications for snippets of her voice, aiming to guard against imitators using AI.